A flood of drug candidates increases the need for novel assays to measure liver toxicity. We have developed methods for quantifying multiple metabolic fluxes in vivo using 2H2O labeling. Here we propose to validate this approach as a toxicity screen. We will test 14 hepatotoxins with diverse mechanisms of action for their ability to alter liver cell turnover (hepatocytes, endothelial cells, total liver cells), mitochondrial biogenesis, and collagen synthesis as markers of liver damage and homeostasis. We will correlate these measurements with changes in static markers of cell division; morphology of sinusoidal endothelial cells; mitochondrial cytochrome c; and collagen deposition, respectively. We hypothesize that toxic concentrations of drugs known to affect the latter phenotypes will strongly modulate fluxes in the underlying metabolic pathways. We will compare static markers and 2H labeling for their ability to detect drug-induced subclinical changes at limiting doses/exposure times and to reveal novel toxic activities. Finally, we will screen for gene expression patterns that predict end organ damage as detected by our test. This work will provide a solid foundation for regulatory approval of our toxicity screen, for extension of our work to other end organs, and for use in humans.